Modular Conceptual Modelling Approach and Software for Sewer Hydraulic Computations

A major challenge in urban water management is the identification of cost-effective and future-proof strategies that can cope with the rapid urbanization and changing environmental conditions. Water quantity modelling forms a key-element in the development of such strategies. Conventional detailed hydrodynamic models are not well suited for use in decision support systems due to several important drawbacks. Therefore, this paper presents a novel and computationally efficient conceptual modelling approach for sewer water quantity simulations. A modular framework is considered that combines well-established model structures with machine learning techniques. This flexible framework ensures that even complex flow dynamics can be emulated accurately. An accompanying software tool was developed to facilitate model configuration. As an example, a full hydrodynamic sewer model of a city in Belgium was transformed into a conceptual model. This model delivered precise results, while the calculation time was 10⁶ times shorter than the detailed model.